Measuring PC power with clamp meter

[Robert Downes]

Out of curiosity (and so I can tell whether the bigger power supplies
are actually necessary next time I built a machine) I would like to
measure how much power my PC is currently using, e.g. during 3D games
that place demand on the 3D card, etc.

I would also like a multimeter for general use (checking power supplies
are still behaving, and so on).

Can a cheap clamp meter do both? I'm under the impression that a clamp
meter for active appliance measurments really needs to be "True RMS"
(because of distortion introduced by noisy drives). But is it possible
to measure power by measuring the voltage on the power cord, then
measuring current at different times? Or would the voltage change
significantly as the demand changed, making current-only calculations
useless?

Can a clamp meter perform the same tasks as a standard digital
multimeter (assuming you have test leads with it)?

Basically, is it possible to closely (within 10% of true power
requirement) measure and calculate appliance power use with only a
standard (well, True RMS) clamp meter, or would the full-whack,
expensive "Clamp-on Power Meter" be required to do this?

Thanks for any advice you can offer.

 

[Walter Harley]

Out of curiosity (and so I can tell whether the bigger power supplies
are actually necessary next time I built a machine) I would like to
measure how much power my PC is currently using, e.g. during 3D games
that place demand on the 3D card, etc.

I would also like a multimeter for general use (checking power supplies
are still behaving, and so on).

Can a cheap clamp meter do both?

You're right that there are some issues with measured versus actual power
consumption for switch-mode power supplies.

But perhaps the biggest problem you'll run into is that the power rating of
the supply has not much to do with how much power it can actually produce;
and it also has not much to do with how much it draws from the power line.

PC power supplies are rated in terms of DC output power: for instance, if a
supply is able to produce 20 amps at 5 volts, plus 3 amps at 12 volts, plus
1 amp at -12 volts, it would be rated at (20*5 + 3*12 + 1*12) = 148W. It
might draw 190W from the power line, because it's not totally efficient.

So, suppose you see your supply drawing 180W from the power line. Should
you conclude you need a 150W supply? (I'm making these numbers up, by the
way; real supplies these days are rated higher than that.) No, because
maybe you're drawing 30 amps at 5v and nothing at all from the 12v supplies.
(Unlikely, but not impossible.)

The most accurate way to rate a PC power supply is to add up the amps of the
various devices inside the PC - each hard drive, the motherboard, each CD,
and so forth - and do the math. Short of simultaneously measuring the
actual OUTPUT current of each power supply line, you're not going to get
much better.

The best way - not the most accurate - is to decide whether you've got an
average, above-average, or below-average system, and buy an average,
above-average, or below-average power supply accordingly. After all, 10
million PC users can't be wrong.

 

[Sharpy]

I would also like a multimeter for general use (checking power supplies

are still behaving, and so on).

Try the following link for a ATX POWER TESTER...
http://www.dansdata.com/quickshot018.htm

Sharpy

----------------------------

 

[Robert Downes]

The best way - not the most accurate - is to decide whether you've got an
average, above-average, or below-average system, and buy an average,
above-average, or below-average power supply accordingly. After all, 10
million PC users can't be wrong.

Ah, damn. I'm already doing that. And I've no idea what capacity UPS
(uninterrupted power supply) I need to put both my machines, and monitor
on a battery backup line.